Finding the Needle in the Haystack: Characterization of the Catalytic and Binding Specificity of Thymine DNA Glycosylase (TDG)

Abstract

Thymine DNA glycosylase (TDG) is a base excision repair enzyme responsible for the repair of G·T mispairs within a CpG sequence context. G·T frequently arises at these sites because of deamination of 5-methylcytosine (5-MeC), which occurs primarily at CpG sites. These mismatches are difficult to locate because they consist of two normally-occurring nucleotides. Our first study examined the mechanisms that TDG employs for catalytic specificity. Using uracil, thymine, and several 5-halouracils (5-FU, 5-ClU, and 5-BrU), we systematically increased the size of the C5 substituent (U<FU<ClU≤T<BrU) while altering the 5' base pair (defining CpG context) and changing the base opposing the target to adenine (analogous to a A:T base pair). The increasing size of C5 substituents results in greater degrees of context dependence in catalysis, likely due to a greater necessity for optimal base-flipping promoted by sequence-specific contacts. Notably, changing the pairing partner of the target base from guanine to adenine results in an approximate 18,000-fold decrease in activity against thymine. Next, the 2:1 binding complex observed in the TDG_core:Ap-DNA crystal opened questions regarding binding stoichiometry and affinity for substrate and undamaged DNA. Using fluorescence anisotropy in equilibrium binding conditions accompanied by EMSAs, we determined that TDG is able to bind mismatches at nanomolar affinities with a second subunit binding at much higher concentrations. We used the non-hydrolysable substrate analogs U<super>F</super> and T<super>F</super> for these experiments to observe formation of the pre-catalytic complex. Using DynaFit, binding affinities were determined for mispairs G∙U<super>F</super> (K_d1= 0.63 ± 0.16 nM) and G∙TF (K_d1= 18 ± 3 nM), abasic DNA (Kd1= 1.4 ± 0.4 nM), an undamaged CpG site (K_d1= 63 ± 10 nM), and undamaged non-specific DNA (K_d1= 293 ± 64 nM). We concluded from these data that TDG is unlikely to bind in a 2:1 conformation in vivo. In another study, we found that N-terminal residues 56-110 significantly contribute to binding affinity and catalysis for damaged and undamaged DNA. The 56-110 region enhances binding of DNA containing G∙U<super>F</super> 64-fold, G∙T<super>F</super> 190-fold, and CpG sites 70-fold. These results extend our understanding of the specificities of the catalytic mechanism of TDG.